Authors: Lav Gupta; Raj Jain; Mohammed Samaka
Addresses: Department of Computer Science and Engineering, Washington University in St. Louis, St. Louis, MO 63130, USA ' Department of Computer Science and Engineering, Washington University in St. Louis, St. Louis, MO 63130, USA ' Department of Computer Science and Engineering, Qatar University, Doha, Qatar
Abstract: Application service providers (ASPs) obtaining resources from multiple clouds have to contend with different management and control platforms employed by the cloud service providers (CSPs) and network service providers (NSPs). Distributing applications on multiple clouds has a number of benefits, but absence of a common multi-cloud management platform that would allow ASPs dynamic and real time control over resources across multiple clouds and interconnecting networks makes this task arduous. Open application delivery network (OpenADN), a multi-cloud management and control platform, fills this gap. However, performance issues of such a complex, distributed and multi-threaded platform, not tackled appropriately, may neutralise some of the gains accruable to the ASPs. In this paper, we establish the need for and methods of collecting precise and fine-grained behavioural data of OpenADN like platforms that can be used to optimise their behaviour to control operational cost, performance (e.g., latency) and energy consumption.
Keywords: software defined infrastructure; application delivery platforms; profiling; inter-cloud; cloud services; network services; application service providers; ASPs; OpenADN; application delivery networks; multi-cloud; distributed systems; optimisation; cloud computing; operational cost; performance; latency; energy consumption.
International Journal of Communication Networks and Distributed Systems, 2016 Vol.17 No.3, pp.296 - 314
Accepted: 02 Jun 2016
Published online: 28 Oct 2016 *